Skip to Main content Skip to Navigation
Journal articles

Regularized Nonlinear Acceleration

Damien Scieur 1 Alexandre d'Aspremont 2 Francis Bach 1
1 SIERRA - Statistical Machine Learning and Parsimony
DI-ENS - Département d'informatique de l'École normale supérieure, CNRS - Centre National de la Recherche Scientifique, Inria de Paris
Abstract : We describe a convergence acceleration technique for generic optimization problems. Our scheme computes estimates of the optimum from a nonlinear average of the iterates produced by any optimization method. The weights in this average are computed via a simple linear system, whose solution can be updated online. This acceleration scheme runs in parallel to the base algorithm, providing improved estimates of the solution on the fly, while the original optimization method is running. Numerical experiments are detailed on classical classification problems.
Document type :
Journal articles
Complete list of metadatas

Cited literature [31 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01384682
Contributor : Alexandre d'Aspremont <>
Submitted on : Wednesday, October 30, 2019 - 6:24:32 PM
Last modification on : Tuesday, September 22, 2020 - 3:57:35 AM

File

1606.04133.pdf
Files produced by the author(s)

Identifiers

Collections

Citation

Damien Scieur, Alexandre d'Aspremont, Francis Bach. Regularized Nonlinear Acceleration. Mathematical Programming, Springer Verlag, 2018, ⟨10.1007/s10107-018-1319-8⟩. ⟨hal-01384682v2⟩

Share

Metrics

Record views

69

Files downloads

245